Part Number: AM6548
In the arm documentation for the R5F, the number of processor cycles for the VMLA command (double precision) is 13 http://infocenter.arm.com/help/index.jsp?topic=/com.arm.doc.ddi0460c/I1019986.html . Is this number correct? For me, this number makes no sense, since the number of clock cycles for VMUL and VADD (each double precision) adds up to 6. Why is there a VMLA instruction on the R5F that consumes more processor cycles than the individual instructions? I always thought such a command was there to save cycles?
